The Daily Receipts

TrendForce expects shortages of high-end MLCCs, the tiny capacitors packed around AI server CPUs, power modules and DDR5 memory, to last through the first half of 2027 as AI inference lifts CPU demand, even though consumer-grade MLCC spot prices slipped about 5% in early October 2026 High-end MLCC shortages are expected to last through the first half of 2027

Microsoft's most recently reported quarterly cash spending on property and equipment reached $39.37 billion

TrendForce expects AI agents to change how servers are built, moving from the traditional one CPU for every eight GPUs or ASICs toward 1:2 or even 1:1, which it forecasts will push server CPU shipments up more than 40% in 2027, after 31% growth in 2026

Even as AI investment surges, US payroll employment in computer systems design and related services, the custom-programming and IT-consulting industry, has kept slipping. The Bureau of Labor Statistics' preliminary September 2026 estimate put it at 2.355 million computer-systems-design jobs, about 32,000 fewer than a year earlier

OpenAI's own researchers are leaning on coding agents at a compounding pace. Epoch AI's analysis of OpenAI's self-reported usage data finds the typical researcher's internal coding-agent usage, priced at API list rates, has been doubling roughly every 34 days

HBM4, the next generation of high-bandwidth memory feeding Nvidia's upcoming GPUs, costs far more to make than the HBM3e it replaces. Fubon Research estimated in a report TrendForce cited in August 2026 that HBM4 for Nvidia GPUs costs $31 to $32 per gigabyte, nearly double HBM3e's $17 to $18